hqq-quantization
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill provides instructions for installing various machine learning packages from official registries (PyPI) and downloading pre-trained models from the Hugging Face Hub.
- [INDIRECT_PROMPT_INJECTION]: The skill documents workflows for loading and serving models that ingest external data (weights and user prompts). This represents a standard ingestion surface for Large Language Models.
- Ingestion points:
AutoModelForCausalLM.from_pretrained(model weights/config from Hugging Face),model.generate(user-supplied text inputs). - Boundary markers: None specified in the provided instructions.
- Capability inventory: Model inference, file saving (
save_pretrained), and network operations for model downloading. - Sanitization: Standard model loading and generation procedures are used without explicit input sanitization described in the skill.
Audit Metadata